https://github.com/rust-lang/rust-clippy/pull/17681 got merged a while ago - this makes it so the `regex_creation_in_loops` clippy `expect` "fails", since it is actually now not triggered.
Unfortunately, this is only in beta from the looks of it at the moment, so I can't just remove it entirely. This change just makes it an unconditional `allow` with a note to get rid of it at some point once it's stable.
Note this only affects macOS, where we use this. This was fine because it's only actually triggered once, and then cached via a `OnceLock`.
Making it so that new features have to go through a feature request first. I'm somewhat not a fan of getting random feature PRs out of the blue - I feel bad saying no too. I would rather be upfront and say that look, if you want a feature, make a feature request and discuss it first before you waste time/tokens/whatever on it.
The way we're doing this for now is:
- If it's less than 100, always show 2 decimal places unless it ends in 0; if so, trim it away
- If it's less than 1000, then do the same but with 1
- If it's over 1000, show no decimal places.
The best way of doing this is probably something like a workspace... but I'm not really in the mood to refactor this into a workspace at the moment, so this will have to do :p
Enable binary disk capacity units with:
```toml
[disk]
use_binary_prefix = true
```
Or pass `--disk_use_binary_prefix` (alias `--disk-use-binary-prefix`). The CLI flag takes precedence over a false config setting. Decimal units remain the default.
Used, Free, and Total use the existing binary conversion helper. For example, 536,870,912,000 bytes changes from `537GB` to `500GiB`. Percentages, sorting by raw byte values, I/O rates, and network units retain their existing behavior.
The change passes the option from argument/config parsing through application settings and collected disk data to cell formatting. Four regression tests cover option precedence and both CLI spellings, collection-to-cell wiring, unit boundaries, missing values, percentages, I/O behavior, and sorting.
This apparently changes the preference directory to use `RoamingAppData` rather than `LocalAppData` (https://codeberg.org/dirs/dirs-rs#7); I don't think this affects me though as I don't find any usage of it.
This adds very basic GPU support for getting process usage for Intel GPUs on Linux. Note this PR does not handle memory/VRAM usage for now. The implementation is basically a clone of the AMD GPU one.
Doing this as refactoring work before I start working on supporting Intel GPUs, which are expected to share the same collection code. This should have no functional change.
This PR makes it so that we should hopefully avoid waking up NVIDIA GPUs while gathering info (on Linux only). It mostly follows how we were doing it for AMD devices, which required a bit of refactoring to get it to work with the NVML library I was using.
Note that, as mentioned above, it does not change behaviour for Windows (or other OSes that work with NVIDIA GPUs) at the moment, as I need to use some fairly OS-specific logic to get it to work on each platform from the looks of it.
Unfortunately, I no longer have any Nvidia-based machines to test with, let alone laptops with Optimus... so this may be hard to verify whether it works on my end.
Pins curl and xz-utils versions for the netbsd pre-build script. This pins it to the most recent versions as of writing.
Note: may be worth exploring just making a custom image that has this rather than doing it each time beforehand.
I had made it so it only used stable, which was generally right but not always (e.g. I downgrade Rust version like right now).
Note this had no effect on actually building the binaries as those use cross.
Clement Tsang